Hollow carbon spheres as ideal supports for hydrogen spillover studies using ruthenium and cobalt nanoparticles

Abstract

Introduction Cobalt Fischer‐Tropsch Synthesis catalysts are generally doped with small amounts of noble metals that serve as reduction promoters to enhance the catalytic activity of the cobalt active sites [1] . This is because the promoter metals are able to dissociate hydrogen gas at a low temperature which then also lowers the reduction temperature of the cobalt oxide to cobalt. This then prevents easy deactivation of the cobalt catalysts that can be induced by high temperature activation. Hydrogen spillover has been invoked to explain the observed effect of these metals as promoters on cobalt catalysts. Two types of hydrogen spillover processes can be envisaged; (1) primary hydrogen spillover, whereby the promoter [i.e., initiator] is in contact with the cobalt oxide [i.e, acceptor] and the dissociated hydrogen atoms can move from the initiator through the direct interface to interact with the acceptor and (2) secondary hydrogen spillover, in this process the initiator and the acceptor materials are separated by some distance and hydrogen spillover can only happen by the dissociation of the hydrogen molecule on the initiator followed by a migration of the atomic hydrogen on a carrier (or catalyst support) to the acceptor material [i.e., cobalt oxide] [2]. Few model catalysts exist that can provide direct evidence of the existence of a type of hydrogen spillover that is dominant on Fischer‐Tropsch like catalysts. In this study mesoporous hollow carbon spheres (MHCS) were used as model supports to study whether both the primary and secondary hydrogen spillover were prominent during catalyst activation and Fischer‐Tropsch synthesis. Experimental MHCS were prepared as shown in Fig 1 (a). Three Co catalysts (15% loading) were prepared (1) Ru@MHCS@Co, with Ru nanoparticles and Co nanoparticles separated by the carbon shell, (2) CoRu/MHCS, Ru and Co co‐precipitated outside MHCS and (3) Co/MHCS, Co outside MHCS. Materials were thoroughly characterized using electron microscopy before being tested under Fischer‐Tropsch conditions at 220 ° C and 10 bar. Results and Discussion Scanning electron microscopy (SEM) analysis of the silica template and hollow carbon spheres gave respective average sizes of 340 nm and 290 nm, thus showing that the silica spheres shrunk as they were heated up to 900 ° C before the carbonization process (Fig 1(b,c) and Fig 2 (a,b)). The resulting hollow carbon spheres retained their spherical nature hence showing no significant breakage of the MHCS. Transmission electron microscopy (TEM) analysis of the materials showed that indeed the spheres were hollow and they had Ru nanoparticles with an average size of 4.1 nm embedded on its walls (Fig 1(e) and Fig 2 (c)). The loaded Co nanoparticles had an average particles size of approximately 5.9 nm on all three catalysts (Fig 1(f) and Fig 2 (d)). MHCS show a distinct roughness under TEM imaging suggesting high porosity of the materials which is necessary to allow reactants to access the encapsulated Ru nanoparticles. TEM tilting over a single axis proved that all the Ru nanoparticles are encapsulated inside the MHCS. Loading of Co nanoparticles outside the MHCS allowed for decoupling of the spillover effects from those that require direct Ru and Co direct contact. Electron Probe Micro‐Analysis (EPMA) large area mapping analysis proved that the metal nanoparticles are well dispersed on the MHCS and thus was ideal materials to study the spillover process (Fig 3). The Fischer‐Tropsch catalytic reaction of the three catalysts was compared and gave a Co time yield in terms of carbon monoxide and hydrogen conversion to hydrocarbons as follows; CoRu/MHCS > Ru@MHCS@Co Co/MHCS. Electron microscopy has therefore helped in following the preparation of a functional material where the promoter effects of Ru using MHCS could be evaluated. I was also observed that a close proximity of Ru and Co nanoparticles was vital for an improved catalytic performance when compared to the case where the Ru and Co nanoparticles were separated by a potential hydrogen transporting material.